Skip to content

[ET-VK] Fix force_fp16 texture bias being silently rejected for CONTIGUOUS_ANY ops#18770

Merged
meta-codesync[bot] merged 2 commits intogh/SS-JIA/516/basefrom
gh/SS-JIA/516/head
Apr 8, 2026
Merged

[ET-VK] Fix force_fp16 texture bias being silently rejected for CONTIGUOUS_ANY ops#18770
meta-codesync[bot] merged 2 commits intogh/SS-JIA/516/basefrom
gh/SS-JIA/516/head

Conversation

@SS-JIA
Copy link
Copy Markdown
Contributor

@SS-JIA SS-JIA commented Apr 8, 2026

Stack from ghstack (oldest at bottom):

The force_fp16 path in TagMemoryMetaPass applies ANY_TEXTURE to
bias ops toward texture storage. However, try_constrain_with_arg_repset
has a packed-dim compatibility check that requires ALL of the source
repset's PDIs to exist in the output repset. ANY_TEXTURE has 3 texture
layouts (WP, HP, CP) but CONTIGUOUS_ANY outputs only support WP, so
the check fails and the texture bias is silently dropped.

Without the bias, buffer storage cascades from ops that must use buffer
(e.g. embedding with vocab exceeding texture limits) into downstream ops
that could use texture, causing unnecessary buffer↔texture transitions.

Fix: check PDI compatibility against the intersection of arg and source
repsets (what would actually be applied) rather than the raw source. The
intersection of ANY_TEXTURE ∩ CONTIGUOUS_ANY = WIDTH_PACKED_TEXTURE,
which IS compatible with the output.

Authored by Claude.

Differential Revision: D100004702

…GUOUS_ANY ops

The `force_fp16` path in `TagMemoryMetaPass` applies `ANY_TEXTURE` to
bias ops toward texture storage. However, `try_constrain_with_arg_repset`
has a packed-dim compatibility check that requires ALL of the source
repset's PDIs to exist in the output repset. `ANY_TEXTURE` has 3 texture
layouts (WP, HP, CP) but `CONTIGUOUS_ANY` outputs only support WP, so
the check fails and the texture bias is silently dropped.

Without the bias, buffer storage cascades from ops that must use buffer
(e.g. embedding with vocab exceeding texture limits) into downstream ops
that could use texture, causing unnecessary buffer↔texture transitions.

Fix: check PDI compatibility against the intersection of arg and source
repsets (what would actually be applied) rather than the raw source. The
intersection of `ANY_TEXTURE ∩ CONTIGUOUS_ANY` = `WIDTH_PACKED_TEXTURE`,
which IS compatible with the output.

Authored by Claude.

Differential Revision: [D100004702](https://our.internmc.facebook.com/intern/diff/D100004702/)

[ghstack-poisoned]
@pytorch-bot
Copy link
Copy Markdown

pytorch-bot Bot commented Apr 8, 2026

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/18770

Note: Links to docs will display an error until the docs builds have been completed.

❌ 1 Cancelled Job, 2 Unrelated Failures

As of commit f8c5861 with merge base 4afd7f9 (image):

CANCELLED JOB - The following job was cancelled. Please retry:

BROKEN TRUNK - The following jobs failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@meta-cla meta-cla Bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Apr 8, 2026
@github-actions
Copy link
Copy Markdown

github-actions Bot commented Apr 8, 2026

This PR needs a release notes: label

If your change should be included in the release notes (i.e. would users of this library care about this change?), please use a label starting with release notes:. This helps us keep track and include your important work in the next release notes.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "release notes: none"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.

…d for CONTIGUOUS_ANY ops"

The `force_fp16` path in `TagMemoryMetaPass` applies `ANY_TEXTURE` to
bias ops toward texture storage. However, `try_constrain_with_arg_repset`
has a packed-dim compatibility check that requires ALL of the source
repset's PDIs to exist in the output repset. `ANY_TEXTURE` has 3 texture
layouts (WP, HP, CP) but `CONTIGUOUS_ANY` outputs only support WP, so
the check fails and the texture bias is silently dropped.

Without the bias, buffer storage cascades from ops that must use buffer
(e.g. embedding with vocab exceeding texture limits) into downstream ops
that could use texture, causing unnecessary buffer↔texture transitions.

Fix: check PDI compatibility against the intersection of arg and source
repsets (what would actually be applied) rather than the raw source. The
intersection of `ANY_TEXTURE ∩ CONTIGUOUS_ANY` = `WIDTH_PACKED_TEXTURE`,
which IS compatible with the output.

Authored by Claude.

Differential Revision: [D100004702](https://our.internmc.facebook.com/intern/diff/D100004702/)

[ghstack-poisoned]
@meta-codesync meta-codesync Bot merged commit 85199dd into gh/SS-JIA/516/base Apr 8, 2026
160 of 164 checks passed
@meta-codesync meta-codesync Bot deleted the gh/SS-JIA/516/head branch April 8, 2026 20:57
@meta-codesync meta-codesync Bot temporarily deployed to cherry-pick-bot April 8, 2026 20:57 Inactive
SS-JIA pushed a commit that referenced this pull request Apr 8, 2026
…GUOUS_ANY ops

Pull Request resolved: #18770

The `force_fp16` path in `TagMemoryMetaPass` applies `ANY_TEXTURE` to
bias ops toward texture storage. However, `try_constrain_with_arg_repset`
has a packed-dim compatibility check that requires ALL of the source
repset's PDIs to exist in the output repset. `ANY_TEXTURE` has 3 texture
layouts (WP, HP, CP) but `CONTIGUOUS_ANY` outputs only support WP, so
the check fails and the texture bias is silently dropped.

Without the bias, buffer storage cascades from ops that must use buffer
(e.g. embedding with vocab exceeding texture limits) into downstream ops
that could use texture, causing unnecessary buffer↔texture transitions.

Fix: check PDI compatibility against the intersection of arg and source
repsets (what would actually be applied) rather than the raw source. The
intersection of `ANY_TEXTURE ∩ CONTIGUOUS_ANY` = `WIDTH_PACKED_TEXTURE`,
which IS compatible with the output.

Authored by Claude.
ghstack-source-id: 364280901
@exported-using-ghexport

Differential Revision: [D100004702](https://our.internmc.facebook.com/intern/diff/D100004702/)
jpiat pushed a commit to jpiat/executorch that referenced this pull request Apr 14, 2026
…GUOUS_ANY ops

Pull Request resolved: pytorch#18770

The `force_fp16` path in `TagMemoryMetaPass` applies `ANY_TEXTURE` to
bias ops toward texture storage. However, `try_constrain_with_arg_repset`
has a packed-dim compatibility check that requires ALL of the source
repset's PDIs to exist in the output repset. `ANY_TEXTURE` has 3 texture
layouts (WP, HP, CP) but `CONTIGUOUS_ANY` outputs only support WP, so
the check fails and the texture bias is silently dropped.

Without the bias, buffer storage cascades from ops that must use buffer
(e.g. embedding with vocab exceeding texture limits) into downstream ops
that could use texture, causing unnecessary buffer↔texture transitions.

Fix: check PDI compatibility against the intersection of arg and source
repsets (what would actually be applied) rather than the raw source. The
intersection of `ANY_TEXTURE ∩ CONTIGUOUS_ANY` = `WIDTH_PACKED_TEXTURE`,
which IS compatible with the output.

Authored by Claude.
ghstack-source-id: 364280901
@exported-using-ghexport

Differential Revision: [D100004702](https://our.internmc.facebook.com/intern/diff/D100004702/)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported meta-exported

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants